Research performed during the past 5 years has provided a considerable amount of evidence to support the contention that the initial interaction of LPS (lipid A) with cells is mediated by distinct plasma membrane proteins. Some of these interactions may be solely involved in removal and eventual degradation of LPS whereas others may play a critical role in transmembrane signaling. Interactions that appear to be limited to a removal function have been assigned to the lipoprotein scavenger receptor or CD18 where R-form LPS, lipid A, or partial lipid A structures such as lipid IVa appear to be the preferred ligands; S-form LPS appears not to interact with these membrane proteins. Whether these interactions reflect events that occur in vivo remains to be definitively established. Moreover, the scavenger receptor and CD18 do not have a role in mediating LPS-induced transmembrane signaling. Photochemical crosslinking studies performed by Morrison and colleagues and by Dziarski (1991a,b) have revealed an LPS-binding membrane protein with an apparent molecular weight 70,000-80,000. This protein binds the lipid A of LPS as well as the carbohydrate backbone of peptidoglycan. Studies with monoclonal antibodies to this protein show that the presence of antibody blocks LPS binding, suggesting that engagement of this protein leads to transmembrane signaling. However, a definitive evaluation of the role of this protein in mediating LPS effects will require complete purification and/or gene cloning. Perhaps the most important advance in our understanding of how LPS acts is derived from the studies of Ulevitch, Tobias, and colleagues wherein the LBP/CD14-dependent pathway of cell stimulation has been identified. This pathway has particular importance for LPS recognition and signaling by cells such as monocytes/macrophages or polymorphonuclear leukocytes that constitutively express CD14. The importance of the LBP/CD14-dependent pathway has been definitively demonstrated by experiments using immunologic, biochemical, and molecular biologic approaches. Available data are consistent with a model for a heterodimeric LPS receptor that consists of CD14 and an as yet unidentified additional protein(s). Clearly a major goal for future research will be to elucidate fully the additional proteins involved in recognition of LPS.